This paper proposes a method for learning embedding representations of style-transfer LoRA models based on their internal weight parameters, capturing transformation characteristics without relying on output examples or metadata. To vectorize LoRA models with multiple layers, we extract internal parameters layer-wise, then flatten and reduce dimensionality, representing each model as a sequence of low-dimensional vectors. Using this vector sequence as input, we perform triplet-based metric learning with a Triplet Network composed of three weight-sharing Transformer encoders and an MLP aggregation module. This framework estimates visual similarity between images transformed by different LoRA models. We conduct automatic evaluation of training validity, verify alignment with human relative similarity judgments, and perform retrieval-based ranking assessment. The results demonstrate that the proposed method learns embeddings consistent with human perception and enables stable retrieval of similar LoRA models.

This paper proposes a retrieval algorithm that ranks LoRA adapters for additional merging, given a target image and the LoRA adapter currently used for image generation. In recent image generation workflows, multiple LoRA adapters are often applied simultaneously. However, selecting appropriate adapters to merge still relies on manual inspection of generated samples and associated metadata. We trained a Transformer-based classification model with token embedding to determine whether a given image is generated by a specific pair of LoRA adapters. This trained model can receive a target image, a fixed LoRA adapter in use, and candidate LoRA adapters, and then output a score reflecting the probability that two adapters, when merged, contribute to generating the target image. By ranking candidates based on this score, the proposed method can retrieve the LoRA adapter to be additionally merged. We implemented a retrieval system supporting 100 LoRA adapters, and both automatic evaluation and a user experiment demonstrate the effectiveness of the proposed approach.

In this study, we evaluate the effect of model merging in ad-hoc retrieval tasks. Model merging is a technique that combines the diverse characteristics of multiple models. We hypothesized that applying model merging to domain-specific ad-hoc retrieval tasks could improve retrieval effectiveness. To verify this hypothesis, we merged the weights of a source retrieval model and a domain-specific (non-retrieval) model using a linear interpolation approach. A key advantage of our approach is that it requires no additional fine-tuning of the models. We conducted two experiments each in the medical and Japanese domains. The first compared the merged model with the source retrieval model, and the second compared it with a LoRA fine-tuned model under both full and limited data settings for model construction. The experimental results indicate that model merging has the potential to produce more effective domain-specific retrieval models than the source retrieval model, and may serve as a practical alternative to LoRA fine-tuning, particularly when only a limited amount of data is available.

In the modern era of digitization, information retrieval has become an indispensable part of daily life. In the future, with further advances in digital technology, information retrieval is expected to become even more diverse and complex. Therefore, we propose an information organization method using a whiteboard and sticky notes in combination with Mixed Reality (MR) devices. We use this method to organize virtual objects displaying information about searched products on a real whiteboard and decide on the items to purchase. With this proposed approach, users can interact with search results in a spatial manner, affixing them to the whiteboard as sticky notes and adding handwritten notes to virtual objects, offering a more intuitive and efficient way to decide on the purchase of products.

Deep hashing has been widely used to approximate nearest-neighbor search for image retrieval tasks. Most of them are trained with image-label pairs without any inter-label relationship, which may not make full use of the real-world data. This paper presents deep hashing, named HA2SH, that leverages multiple types of labels with hierarchical structures that an ethnological museum assigns to their artifacts. We experimentally prove that HA2SH can learn to generate hashes that give a better retrieval performance. Our code is available at https://github.com/wbw520/minpaku.

This paper introduces a new test collection for ad-hoc dataset retrieval, which have been developed through a shared task called Data Search in the fifteenth NTCIR. This test collection consists of dataset collections derived from the US and Japanese governments' open data sites (i.e., Data.gov and e-Stat), as well as English and Japanese topics for these collections. Organizing the shared task in NTCIR, we conducted relevance judgments for datasets retrieved by 74 search systems, and included them in the test collection. In addition to the detailed description of the test collection, we conducted in-depth analysis on the test collection, and revealed (1) what techniques were used and effective, (2) what topics were difficult, and (3) large topic variability in the dataset retrieval task.

In this study, we developed an Internet of Things (IoT) monitoring device to monitor over the people inside a room. We collected sensor data at a specific location using the device. Based on the data, we tried to predict the behavior of the person at that location. Monitoring and predicting human daily behavior is trivial task. Most of the research on monitoring and predicting daily life behavior are based on the data available from smart home [7] [16] [17]. But smart home is expensive compare to normal home, as different kind of sensor are attached in the room and have more facilities. So, we developed a low cost IoT monitoring device and predict the daily life behavior of human from the sensor data taken from the device. We can extract information from the daily life behavior and share it with the family living in distant places.

In this paper, we proposed a method for determining whether a given restaurant review comment is a repeater's review, or not. We often use restaurant review sites to decide which restaurant to go to. When we read a restaurant review comment, we can know whether the reviewer is a repeater of the restaurant. If a certain restaurant has many repeaters, the restaurant must be great. However, restaurant review sites usually do not provide a "revisit rate". Therefore, we tackle a problem for determining whether a review is a repeater's review, or not. There are many sentences in a review comment that are completely not useful for determining whether the review is a repeater review, such as what was ordered, what was delicious, or how was the price. To confront such difficulties, we have taken the following approach. First, very various features are extracted from review comments so as not to miss the features that represent repeaters' reviews. Next, from the very various features, only the necessary features that really contribute to the classification is selected by a feature selection method. Finally, classification is performed using a classifier. We have implemented the proposed method using super-CWC [12], a state-of-the-art feature selection method, and SVM. The experimental results show that the proposed method is better than other methods.

In this study, we propose a method to classify genre-based films using film screenplays. The proposed method vectorizes films into two aspects, i.e., film content (i.e., what films tell viewers) and screenplay structure (i.e., how the films narrate stories), and classifies film genres using the support vector machine method. We applied the Doc2Vec algorithm to screenplay structure and to handle film content. In film production, for vectorizing films, we used the statistics of the four screenplay elements: scene, action, dialogue, and transition. Compared with baseline methods, the evaluation showed that the proposed method is better for classifying films of specific genres.

In this paper, given the subject-verb pairs, we propose a computational model to express the difference of the meaning of a verb when the subject has changed. We propose a method to generate metaphorical expressions consist of subject-verb pairs from the model. "Airship swims" is one of the example. It is the expression about the event that an airship flies in the sky gracefully. There are a few reasons why the expression is accepted for people. "airship flie" and the motion of a sea creature, for example "whale swims" represent both "the normal move in a space" and there is a similarity. Given the input ("airship," "fly"), we propose a method to detect a verb "swims" to generate metaphorical expressions considering these similarity. At first, we test which vectorization method is the best as the vectorization of a subject-verb pair. We calculate a transformation matrix to conserve between the meaning of (non-human subject, verb) pairs and the meaning of ("man", verb) pairs. We calcurate the transformation matrix between them using the stable meaning verbs as the anchors. In this paper, we test an hypothesis that we can use these transformation matrices to find an appropriate verb considering the difference of the meaning occured from the subjects. We gather 67 cases of target figurative expressions from Web. We evaluated the proposed method by defining the information retrieval problem of verbs.

This study proposes a method for retrieving and ranking posts from social network services(SNSs) by specifying and providing feedback on the context of posts. Current search systems for SNS posts cannot handle user intent with regard to the context of posts to be retrieved, mainly owing to the incompleteness of SNS posts, i.e., they do not contain the users' contexts (e.g., situations or preferences) of users posting messages. Hence, we propose a search method that accepts two kinds of queries, namely, content queries and context queries, and that updates these queries based on the user feedback with special attention to the contexts of posts. Our search method considers the whole SNS dataset as a graph and the nodes surrounding each post as its context; to find relevant posts in terms of content and context, our method propagates user feedback via this graph. Our experimental results based on a Twitter test collection revealed that our proposed method showed improved retrieval performance as compared with conventional SNS retrieval and relevance feedback. In addition, we could detect the optimal parameters for feedback propagating.

A recent study on information refinding reported that 44% of Web page visits and 33% of Web queries involved revisiting previously browsed pages. We propose methods for finding previously browsed pages regarded as coordinate pages of currently browsed pages. Intuitively, the notion of coordinate pages means that both of them belong to an identical class. To find the coordinate pages for given pages, we use a user's browsing and search behavior, such as her query log and tab usage, as well as link navigation. Our page revisiting methods were implemented within a Web browser, so that users can find those previously browsed pages while browsing and searching. We conducted experiments in which our methods outperformed conventional baseline methods in terms of page revisiting.

The effectiveness of retrieval decreases with the increase in query length. We target at sentential queries and propose a method for improving their retrieval performance, called query rewriting. Briefly, given a sentential query, our method acquires paraphrases from the noisy Web and uses them to avoid returning no answers. In particular, since a relation can be represented either intensionally (referred to as paraphrase templates) or extensionally (referred to as coordinate tuples), the mutual reinforcement between them are taken into account. The experimental results show that for declarative sentences, the average precision of our method is 68.1%, compared to 44.2% of the baseline. Besides, the relative recall of our method is 95.9%, nearly 3 times compared to that of the baseline. While for questions, the average precision of our method is 46.9%, compared to 39.9% of the baseline. We also show the effectiveness of query rewriting in two applications.

This study investigated query formulations by users with {\it Cognitive Search Intents} (CSIs), which are users' needs for the cognitive characteristics of documents to be retrieved, {\em e.g. comprehensibility, subjectivity, and concreteness. Our four main contributions are summarized as follows (i) we proposed an example-based method of specifying search intents to observe query formulations by users without biasing them by presenting a verbalized task description;(ii) we conducted a questionnaire-based user study and found that about half our subjects did not input any keywords representing CSIs, even though they were conscious of CSIs;(iii) our user study also revealed that over 50\% of subjects occasionally had experiences with searches with CSIs while our evaluations demonstrated that the performance of a current Web search engine was much lower when we not only considered users' topical search intents but also CSIs; and (iv) we demonstrated that a machine-learning-based query expansion could improve the performances for some types of CSIs.Our findings suggest users over-adapt to current Web search engines,and create opportunities to estimate CSIs with non-verbal user input.

We introduce the problem of domain adaptation for content-based retrieval and propose a domain adaptation method based on relative aggregation points (RAPs). Content-based retrieval including image retrieval and spoken document retrieval enables a user to input examples as a query, and retrieves relevant data based on the similarity to the examples. However, input examples and relevant data can be dissimilar, especially when domains from which the user selects examples and from which the system retrieves data are different. In content-based geographic object retrieval, for example, suppose that a user who lives in Beijing visits Kyoto, Japan, and wants to search for relatively inexpensive restaurants serving popular local dishes by means of a content-based retrieval system. Since such restaurants in Beijing and Kyoto are dissimilar due to the difference in the average cost and areas' popular dishes, it is difficult to find relevant restaurants in Kyoto based on examples selected in Beijing. We propose a solution for this problem by assuming that RAPs in different domains correspond, which may be dissimilar but play the same role. A RAP is defined as the expectation of instances in a domain that are classified into a certain class, e.g. the most expensive restaurant, average restaurant, and restaurant serving the most popular dishes. Our proposed method constructs a new feature space based on RAPs estimated in each domain and bridges the domain difference for improving content-based retrieval in heterogeneous domains. To verify the effectiveness of our proposed method, we evaluated various methods with a test collection developed for content-based geographic object retrieval. Experimental results show that our proposed method achieved significant improvements over baseline methods. Moreover, we observed that the search performance of content-based retrieval in heterogeneous domains was significantly lower than that in homogeneous domains. This finding suggests that relevant data for the same search intent depend on the search context, that is, the location where the user searches and the domain from which the system retrieves data.

We propose a query-by-example geographic object search method for users that do not know well about the place they are in. Geographic objects, such as restaurants, are often retrieved using an attribute-based or keyword query. These queries, however, are difficult to use for users that have little knowledge on the place where they want to search. The proposed query-by-example method allows users to query by selecting examples in familiar places for retrieving objects in unfamiliar places. One of the challenges is to predict an effective distance metric, which varies for individuals. Another challenge is to calculate the distance between objects in heterogeneous domains considering the feature gap between them, for example, restaurants in Japan and China. Our proposed method is used to robustly estimate the distance metric by amplifying the difference between selected and non-selected examples. By using the distance metric, each object in a familiar domain is evenly assigned to one in an unfamiliar domain to eliminate the difference between those domains. We developed a restaurant search using data obtained from a Japanese restaurant Web guide to evaluate our method.

We describe methods to search with a query by example in a known domain for information in an unknown domain by exploiting Web search engines. Relational search is an effective way to obtain information in an unknown field for users. For example, if an Apple user searches for Microsoft products, similar Apple products are important clues for the search. Even if the user does not know keywords to search for specific Microsoft products, the relational search returns a product name by querying simply an example of Apple products. More specifically, given a tuple containing three terms, such as (Apple, iPod, Microsoft), the term Zune can be extracted from the Web search results, where Apple is to iPod what Microsoft is to Zune. As a previously proposed relational search requires a huge text corpus to be downloaded from the Web, the results are not up-to-date and the corpus has a high construction cost. We introduce methods for relational search by using Web search indices. We consider methods based on term co-occurrence, on lexico-syntactic patterns, and on combinations of the two approaches. Our experimental results showed that the combination methods got the highest precision, and clarified the characteristics of the methods.

A method is described for discovering coordinate terms, such as "Honda'' and "Nissan,'' for a given term, such as "Toyota,'' as well as their common topic terms, from the query logs of a Web search engine. Coordinate terms are good candidates for use in making comparisons. A HITS-based algorithm is applied to a bipartite graph between coordinate term candidates and co-occurrence patterns to identify coordinate and topic terms. Spectral analysis is used to distinguish coordinate terms corresponding to different aspects of the search term. As a result, we can discover terms related to the terms in a search engine query that reflect the needs and interests of the user.