University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 [email protected] CSA4080:...

52
1 of 52 [email protected] University of Malta CSA4080: Topic 5 © 2004- Chris Staff CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department of Computer Science & AI University of Malta Topic 5: Recommendation Techniques

Transcript of University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 [email protected] CSA4080:...

Page 1: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

1 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

CSA4080:Adaptive Hypertext Systems II

Dr. Christopher StaffDepartment of Computer Science & AI

University of Malta

Topic 5: Recommendation Techniques

Page 2: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

2 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Aims and Objectives

• Global Reconnaissance Techniques– PowerScout– Watson– HyperContext

• Recommender Systems– Amazon– IMDB

Page 3: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

3 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Aims and Objectives

• User Modelling in IR

• User Modelling in Recommender Systems

Page 4: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

4 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Readings

• recommender p36-soboroff.pdf

• SOTA Recommender systems Lit Review.pdf (Chapter 8 - )

• recommender 0329_050103.pdf

Page 5: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

5 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

What is Recommendation?

• Recommendations are suggestions

• It could be a suggestion to watch a particular movie, or to buy a particular product, visit a restaurant (not fish!)

• In hyperspace, this could be a suggestion to follow a path leading to a relevant document, or to visit a document directly

Page 6: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

6 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

What is Recommendation?

• If the recommendation is to do with guidance, then this is related to adaptive navigation

• If the recommendation is based mainly on recommending products, then it is a recommender system

• The two are, or can be, closely related, but the literature tends to deal with them separately

Page 7: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

7 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Examples...

• Global Reconnaissance, Guidance, Personal Information Management Assistants...

• As you browse a user model of your interests is automatically built

• Paths are recommended, or other documents are collected for your perusal

• Usually use IR systems to index, search for, and retrieve relevant documents

Page 8: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

8 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Global Reconnaissance

• PowerScout (Lieberman, 2001)– Automatically builds user model from recently

viewed pages, but based on user’s long-term interaction

– Searches for relevant documents via 3rd party search engine

– Organises results by “Concept”

Why-Surf-Alone.pdf

Page 9: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

9 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Global Reconnaissance

• Watson (Budzik et al, 1998)– Observes user interacting with several

application to build model of user’s information goal

– Anticipates that user is interested in documents similar to ones seen in recent past

– Searches for documents (via 3rd party search engine) and presents list to user

– Short-term user model, with long-term supportbudzik99watson.pdf

Page 10: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

10 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Global Reconnaissance

• HyperContext (Staff, 2000)– Uses Adaptive Information Discovery (AID)

techniques to find remote but relevant information

– Short-term UM, with long-term UM support

HCTCh5.pdf

Page 11: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

11 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

More examples...

• Recommender systems– Content recommendation– Collaborative recommendation

Page 12: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

12 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Recommender Systems

• “What did you think about...?” “Did you like...?”

• Make recommendation based on past experience

• Real world examples: food critic, movie critic, book/novel critic, lecture course critic :-)

Page 13: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

13 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Recommender Systems

• How do you know you can trust somebody’s recommendation?– Because experience has taught you?– Because critic is trusted source of info?– Because a friend/expert likes movies/novels/

food you like?– ???

Page 14: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

14 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Recommender Systems

• Generally two types of recommender system:– Content-based recommendation– Collaborative recommendation

• burke-umuai02.pdf

• recommender 0329_050103.pdf

Page 15: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

15 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Recommender SystemsCollaborative Recommendation

• Usually, ratings-based feedback

• Users must indicate degree to which they like product, product is fit for purpose, etc

• The recommendation is based on the weighted average utility of the product...

• ... of users with the same preferences!– preferences may also include demographics

Page 16: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

16 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Recommender SystemsCollaborative Recommendation

• Do you want recommendations based on all users?

• Or do you want recommendations from other people like you, with your tastes and preferences?

• How can the system work out what you like/prefer/want?– Comparing interactions (purchases, queries,

movies seen, etc.) and identifying trends

Page 17: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

17 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Recommender SystemsCold-Start Problem

• Collaborative recommender systems suffer from the cold start problem

• How do you recommend a new product with no ratings?

• How do you recommend to a new user?

• Content-based recommendation overcomes some problems

Page 18: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

18 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Recommender SystemsContent-based

• Instead of using ratings, use product features

• Identify features using eg., kdd96_quest.pdf– On what basis can products be compared?

Genre, cost, dimensions, etc.• Recommendations can be based on user-

selected feature sets, or on prior interactions– Latter works for frequent recommendations of

similar product (e.g., movie) but not infrequent ones, e.g., camera purchase

Page 19: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

19 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Recommender SystemsCold-Start Problem Revisited

• If user categorisation is automatic (i.e., System believes user U belongs to group G based on past interactions) then cold-start problem for new users

• New products are ok, though, because they will be recommended based on feature similarity

• If user drives feature selection, then is system user-adaptive?

Page 20: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

20 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Recommender Systems

• Both collaborative and content-based recommendation utilise clustering techniques to identify patterns in users and/or products/items

• Most common technique is the Vector Space Model (Topic 6)

• Other IR techniques also used

Page 21: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

21 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in IR andRecommender Systems

• User model is usually created and maintained for information retrieval and recommender systems

Page 22: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

22 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling

• In pure IR, user interaction is usually geared towards selecting relevant documents from a collection/repository

Page 23: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

23 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling

• Is there a user model, even a simple one, in this model of IR?

• If there is, is there a point at which adaptation might be said to take place?

• More next topic...

Page 24: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

24 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in IR

• This part based heavily on www.scils.rutgers.edu/~belkin/um97oh/

Page 25: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

25 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in IR

• In early IR (before automation!) human mediators (e.g., librarians) construct queries on behalf of users – See also, evaluation of boolean model (p289-

blair.pdf)– Search intermediaries still used in some Web-

based question-answering systems, e.g., AskJeeves

Page 26: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

26 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in IR

• As query specification languages became complex (1950s/60s) intermediaries needed to construct queries

• It became useful in systems like SDI to store representations of users’ long-term interests so that new information objects could be routed to them

Page 27: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

27 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in IR

• Initially, user profiles were changed manually on basis of user’s evaluation of search results

• Eventually, SDI could automatically modify profiles based on relevance judgements

• This line of IR developed into information filtering (routing)

Page 28: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

28 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in IR

• Ad hoc IR assumes that information need is just one-time– there is just one information seeking episode– a single query is compared to a static document

collection

• If there is a subsequent query that is submitted by the same user and that is related to a prior query, it is treated as a new episode

Page 29: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

29 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in IR

• In ad hoc IR user may need support to:– Reformulate the query to get better results– Provide relevance feedback so that system can

modify the query (Rocchio, 1966)

• In “queryless” IR (Oddy, 1977) the user need not specify the information need:– user evaluates/rates features of retrieved info– system builds model of user’s interests

Page 30: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

30 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in IR

• ASK-based IR (Belkin et al, 1982)– elicits and represents user’s Anomalous State of

Knowledge rather than specific info need– Associative network represents ASK– Uses rules to compare ASK with document

representations– User ratings of features can auto update ASK

Page 31: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

31 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in IR

• Modelling user goals (Vickery, Vickery & Brooks, 1980s)– to determine the comparison techniques to

apply for different users– users direct elicitation + implication from user

behaviour– long term modelling of user preferences and

“typical” info problems

Page 32: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

32 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in IR

• Models for identifying UM functions in IR– Abstract analysis of IR task. To identify:

• goals of IR

• problems in achieving goals

• what’s necessary for other actors in the system to know of user to achieve goals/overcome problems

query as specification as modelling function

Page 33: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

33 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in IR

– IR interaction as dialogue• what is needed to experience effective conversation

(e.g., Grice’s rules of conversational implicature)

• how can these be modelling in an IR interaction?

models of understanding that each actor has of the other (“I believe that you believe...”, and see Kobsa’s BGP-MS)

Page 34: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

34 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in IR

– Observing user behaviour in IR systems settings

• cognitive task analysis

• failure analysis

• thinking aloud, etc.

Stereotypical models of experience, expertise, search behaviours, “needs”

Page 35: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

35 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in IR

• Overall goal (not Belkin’s words!)– Intelligent agents that can understand user

needs/goals/tasks by observing user behaviour and that can find, retrieve, or even accomplish, what the user had set out to do, without the user necessarily expressing his or her intentions

Page 36: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

36 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in Recommender Systems

• Recommender systems – Content-based (very similar to IR)– Collaborative

• Aim is to make recommendations based on what other, similar, users liked or did

recommender 0329_050103.pdf

Page 37: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

37 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in RS

• In general, let C be the set of all users, and let S be the set of all recommendable items (CDs, books, movies, holidays, documents...)

• Let u be a utility function which measures the usefulness of item s to user c

u:C x S Rwhere R is a totally ordered set (of, e.g., reals)

Page 38: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

38 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in RS

• In RS, utility of an item to a user is usually represented as a rating, how much a particular user liked the item, but it can be any function

• On what basis do we decide that two users are similar?

Page 39: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

39 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in RS

• What information is retained about users?– Demographic information– Interaction history– Ratings given to items

Page 40: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

40 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in RS

• Two main types of algorithm– Memory-based– Model-based

Page 41: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

41 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in RS

• Memory-based algorithm– heuristics that make rating predictions based

on entire collection of previously rated items by users

• Predict rating for user c on item s assuming user has not previously seen item (simplest)

rc,s =1

Nr ′ c ,s

′ c ∈ ˆ C

∑where C is set of N users c’ that aremost similar to user c and who have rateditem s

^

Page 42: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

42 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in RS

• Problem with simplest algorithm...– Doesn’t take into account similarity between

users, only similarity between prior ratings

– sim(c’,c) is the similarity (distance measure)

between two users, k is a normalising function €

rc,s = k sim(c, ′ c ) × r ′ c ,s

′ c ∈ ˆ C

k =1 sim(c, ′ c )′ c ∈ ˆ C

Page 43: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

43 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in RS

• Many ways of deriving user similarity measure

• Normally based on the set of items, Sxy, that both users, x and y, have rated

• Two popular approaches– Cosine-based– Correlation-based

Page 44: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

44 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in RS

• Correlation-based approach

where rx is the average rating given by user x€

sim(x, y) =

(rx,s − r x )(ry,s − r y )s∈Sxy

(rx,s − r x )2 (ry,s − r y )2

s∈Sxy

∑s∈Sxy

_

Page 45: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

45 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in RS

• Cosine-based approach– 2 users x and y are treated as vectors in m-

dimensional space, where m is the number of items in Sxy

sim(x,y) = cos(r x ,

r y ) =

r x •

r y

r x

r y

2

=

rx,sry,s

s∈Sxy

rx,s2

s∈Sxy

∑ ry,s2

s∈Sxy

Page 46: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

46 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in RS

• Memory-based approaches need many ratings to work well

• Default voting improves rating prediction accuracy

Page 47: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

47 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in RS

• Model-based algorithm to measure user similarity– uses collection of ratings to learn a model

which is then used to make rating predictions

– the probability that user c will give a particular rating to item s given that user’s ratings of the previously rated items (Breese et al, 1998).€

rc,s = E(rc,s) = i × Pr(rc,s = i rc, ′ s , ′ s ∈ Sc )i= 0

n

Page 48: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

48 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

User Modelling in RS

• Breese et al proposed two alternative probabilistic models to estimate the probability expression– Cluster model (Naive baysian)

• Users are clustered into groups

– Baysian networks• Each item is a node in the network, with states of

each node represent possible rating values• Network and conditional probabilities are learned

from data

Page 49: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

49 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Collaborative System Shortcomings

• New user problem

• New item problem

• Sparsity– Can initially be resolved using demographic

data

Page 50: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

50 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Conclusion

• IR has users with both long- and short-term interests

• RS has users with mainly long-term interests, although recommendations may be made to users with short-term interests– In which case, the method of interaction is

usually different, and recommendations are based on content

Page 51: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

51 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Conclusion

• In IR, an explicit user model is maintained for long-term support, but a query is a reasonable ad hoc model of the user’s interest

• In RS, users need to be distinguished in the collaborative model, but not in the content model

Page 52: University of Malta CSA4080: Topic 5 © 2004- Chris Staff 1 of 52 cstaff@cs.um.edu.mt CSA4080: Adaptive Hypertext Systems II Dr. Christopher Staff Department.

52 of [email protected] University of Malta

CSA4080: Topic 5© 2004- Chris Staff

Conclusion

• In the next topic we will look at IR models and techniques– Vector-based model– Probabilistic model– Relevance Feedback– Query Reformulation

• We will also look at knowledge and domain representation